Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS5513350 A
Publication typeGrant
Application numberUS 08/250,205
Publication dateApr 30, 1996
Filing dateMay 27, 1994
Priority dateMay 27, 1994
Fee statusLapsed
Also published asCA2146690A1, EP0689136A1
Publication number08250205, 250205, US 5513350 A, US 5513350A, US-A-5513350, US5513350 A, US5513350A
InventorsTimothy G. Griffin, Howard W. Trickey
Original AssigneeAt&T Corp.
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Update constraints in transactions which may abort
US 5513350 A
Abstract
A technique which may be used in a transaction which updates a data set when the data set is subject to constraints and includes operations which may abort. The technique produces an update constraint which may be tested before the data set is altered to determine whether the constraints will be satisfied if the data set is altered. The update constraint of the technique indicates whether the constraints would be satisfied if the transaction could not abort. The update constraint is checked at the application level, while the constraints which cause an operation to abort are checked at the system level. An application of the update constraint to a relational data base embedded in a telephone switch is disclosed.
Images(2)
Previous page
Next page
Claims(10)
What is claimed is:
1. Apparatus for generating code for a transaction on a set of data, the transaction being subject to constraints and the apparatus comprising:
a transaction description which describes operations performed in the transaction, the operations including at least one operation which may abort without affecting the set of data;
a constraint specification which specifies the constraints to which the transaction is subject; and
code generating means responsive to the transaction description and the constraint specification for generating the code, the code generating means including
means for generating code for an update constraint which indicates whether the constraints would be satisfied if the transaction could not abort and which is executed prior to any operation which may abort.
2. The apparatus set forth in claim 1 wherein:
the system in which the code is executed has an application level and a system level;
the code for the update constraint runs at the application level; and
the operation aborts at the system level.
3. The apparatus set forth in either claim 1 or 2 wherein:
the means for generating code generates code for an update constraint which is a weakest precondition.
4. The apparatus set forth in either claim 1 or 2 wherein:
the transaction is an update transaction which alters the set of data.
5. The apparatus set forth in either claim 1 or 2 wherein:
the apparatus is implemented in a data base system; and
the transaction is an update transaction of the data base system.
6. A method of making apparatus for performing a transaction on a data set, the transaction being subject to constraints and involving an operation which may abort,
the method comprising the steps of:
computing from the constraints and a description of the transaction which describes Operations of the transaction an update constraint which indicates whether the constraints would be satisfied if the transaction could not abort; and
generating code which is executed during the transaction and which determines whether the update constraint is satisfied.
7. The method set forth in claim 6 wherein:
the operation which may abort aborts at a system level; and
the step of generating code generates code which is executed at the user level.
8. The method set forth in either claim 6 or claim 7 wherein:
the step of generating code generates code for an update constraint which is a weakest precondition.
9. A method of performing a transaction on a set of data which alters the set of data, the transaction being subject to constraints, the method including the step of altering the data, and the step of altering the data involving an operation which may abort,
the method having the improvement comprising the step of:
prior to performing the step of altering the set of data, testing an update constraint which, when satisfied, indicates whether the constraints would be satisfied if the transaction could not abort, and not altering the set of data if the update constraint is not satisfied.
10. A data base system for performing transactions on a set of data, the transactions being subject to constraints and the data base system including means for updating data, the means for updating data performing an operation which may abort without affecting the set of data, and the database system having the improvement comprising:
means for ensuring integrity of the set of data by testing an update constraint which, when satisfied, indicates that the constraints would be satisfied if the operation could not abort and indicating to the means for updating whether the update constraint is satisfied,
the means for updating altering the set of data only if the update constraint is satisfied.
Description
1 BACKGROUND OF THE INVENTION

1.1 Field of the Invention

The invention concerns systems in general in which operation of the system is subject to constraints and particularly concerns constraints involved in maintaining the integrity of data in a data base system.

1.2 Description of the Prior Art

An important operation in data base systems is changing the data in the system. Operations which change data may involve adding data, deleting data, or updating data, which is a combination of deleting and adding. In the following, operations which change the data base are termed generally update operations.

Care must be taken in an update operation that the data being added to the data base system does not compromise the integrity of the data in the data base system. For example, a data base of employee information may have a record for each employee which includes fields specifying the employee's grade and pay. Pay may be coupled to the grade, so that it must be above a minimum and below a maximum specified for the grade. In such a data base, maintaining the integrity of the data base system requires that when a new employee record is added or an existing record is altered, the data base system check whether the pay in the new or altered record is within the range specified for the grade. A condition which must be met by data which is to be included in a data base in order to maintain the integrity of the data base is termed a constraint, and in many data base systems, any update operation includes a phase called constraint checking, in which the data base system determines whether the constraints required for the operation are satisfied.

One way of doing constraint checking is to do the operation, check any constraints, and if a constraint has been violated, undo the update. However, it is usually possible to determine whether the constraints will be violated before the update is made and do the update only if there is no violation.

Of course, constraint checking takes time, so it is important to do as little as possible of it. In many cases, constraints are dependent on other constraints in such a fashion that only a few of the constraints actually need to be checked. A smaller set of constraints which can be checked with the same results as checking all of the constraints is termed an update constraint. One way of finding an update constraint is to calculate a logical formula called the weakest precondition of the entire set of constraints with regard to the update. The weakest precondition sets forth a set of constraints which must be true if the entire set of constraints is to be true after the update has been performed. See C. A. R. Hoare, "An axiomatic basis for computer programming," in: Communications of the ACM 12(10), October 1969.

However, even testing the constraints of the weakest precondition may involve more work than necessary. As set forth in X. Qian, "An effective method for integrity constraint simplification", in: Fourth International Conference on Data Engineering, 1988, it is possible to take advantage of the fact that the constraints are known to hold in advance. Since that is the case, the weakest precondition can be factored into a first portion of the weakest precondition which is implied by the constraints and a second portion which is not. The AND of these portions is logically equivalent to the weakest pre-condition, and consequently, it possible to check the weakest precondition by checking the second portion only, and the second portion by itself can thus function as the update constraint.

A problem with the foregoing work on update constraints involving the weakest precondition and Qian's further simplification thereof is that it does not take into account the fact that in practical data base systems, certain of the data base system operations which are executed in the course of an update operation may themselves abort, that is, they may terminate without affecting the data base if certain constraints imposed by the data base system are not satisfied. For example, the data base system insert operation which actually inserts a new record into a data base requires that the new record be in fact new; it determines this by checking whether a key which identifies the record to the date base system is the same as that of any existing record in the data base; if it is, the insert operation aborts. The constraints which these operations must satisfy are different from those required to keep the content of the data consistent, and it has heretofore not been apparent how to incorporate such constraints into update constraints.

2 SUMMARY OF THE INVENTION

The inventors have discovered that all that is necessary to solve the problem of update constraints in systems which have operations which may abort is to determine the update constraint using the assumption that the transaction cannot abort. In one aspect of the invention, the discovery is used in methods and apparatus for generating code for transactions which incorporates update constraints determined as describe above. In another, such update constraints are used in methods and apparatus for performing update transactions. An advantage of such update constraints is that the checks for the constraints which cause the system to abort the operation are performed at the system level, while those for the update constraint are performed at the application level.

Other objects and advantages of the apparatus and methods disclosed herein will be apparent to those of ordinary skill in the art upon perusal of the following Drawing and Detailed Description, wherein:

3 BRIEF DESCRIPTION OF THE DRAWING

FIG. 1 is an overview of a data base system in which the invention may be employed;

FIG. 2 shows two relations employed in an example;

FIG. 3 shows constraints employed in the example;

FIG. 4 is a block diagram of apparatus for generating update code according to the invention; and

FIG. 5 is an example of update code generated for the example.

Reference numbers in the Drawing have two parts: the two least-significant digits are the number of an item in a figure; the remaining digits are the number of the figure in which the item first appears. Thus, an item with the reference number 201 first appears in FIG. 2.

4 DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT

The following discussion will begin with an example of the invention; thereupon, a precise logical definition of the invention will be presented.

4.1 Overview of a Data Base System: FIG. 1

FIG. 1 shows a typical data base system 101. The hardware components of system 101 are a processor 111 which can execute computer programs, mass storage device 119, on which data is stored, and user interface 103, which includes a display, a keyboard, and a pointing device such as a mouse.

System 101 results from the execution of the code for a data base management system 113 on processor 111. Data base management system 113 organizes the data in mass storage 119 into a set of relations. Each relation is a table which has columns and rows. The columns define classes of data (for example, employee name, employee department, and pay), while the rows contain related values of the data, for instance a specific employee name and the specific values for employee department and pay which go with that employee name. There are two kinds of relations in data base system 101: base relations 121, in which the data values are actually stored in mass storage 119, and views, that is, relations which do not really exist in mass storage 119 but are constructed on the fly from data in base relations 121.

Data base management system 113 manipulates the data in response to transaction specifications 105 which specify transactions on the data in the data base system. The transaction specifications are provided by users of the data base system. The transaction specifications, specify operations on the relations such as reading data from a relation or writing data to the relation. Data base management system 113 responds to a transaction specification 105 by invoking data base management system primitive routines 116 as needed to perform the operations 117 on mass storage device 119 required to execute the transaction; if the transaction succeeds, the result is returned to the user interface, as indicated by arrow 107; otherwise, an error message 109 is returned.

As pointed out in the Description of the Prior Art, one important class of operations specified by transaction specifiers is update operations, in which the data in relations 121(a) is altered. The part of data base management system 115 which performs update operations appears in FIG. 1 as update code 115. As also pointed out in the Description of the Prior Art, when a transaction involves an update of data in relations 121, data base management system 113 must ensure that the update does not degrade the integrity of data base system 101, and consequently, part of the task of update code 115 is to check whether update constraints have been violated. If they have been, update code 115 does not change the data in relations 121, but instead returns an error message 109. Transactions involving updates are termed in the following update transactions.

4.2 An Example Update Transaction: FIGS. 3-5

In the following, a simple example will be presented to show how weakest preconditions may be used to do constraint checking in data base system 101. The example involves two relations 121. The relations are shown in FIG. 2. The first relation is employees relation 201, which has a row 203 for every employee. The columns of the relation are name 205, for which the values are employee names, sal 207, for which the values are employee salaries, and dept, for which the values are department numbers. Thus, each row 203 specifies an employee's name, his salary, and his department number. As indicated by the underline, the name field functions as the key for each record, that is, the record is identified to data base management system 113 by the value of name.

The second relation is departments 211, which has a row 213 for each department in the business. The columns are dept 215, whose value is the department number, and manager 217, whose value is the name of the manager. The key in departments 211 is dept 215. Departments 211 and employees 201 are related to each other by way of the keys. For example, the manager whose name appears in manager column 217 for a given row 213 of departments 211 also has an entry under his name in employees 201. Because the key of one relation is used as data in the other, the data base management system can construct views which contain information from both relation 211 and relation 201. One example of such a view is a relation which has an entry for each manager and which specifics the manager's name, department, and salary. In making the view, data base management system 113 retrieves the manager's name and department from departments 211 and uses the manager's name in employees 201 to retrieve the value of the manager's salary from column 207.

Even in the two relations 211 and 201, there are a number of constraints on the values of the rows. FIG. 3 shows these constraints 303, expressed in the notation of logic. Constraint 303 states that for all employees in relation 201, the salary must range between two values specified by rain and max. Constraint 305 specifics that for all employees, the department number in column 209 must be one of the department numbers in departments 211. Constraint 307 says that for every row in departments 211, there must be a row in employees 201 such that the value of manager equals the value of name and the value of dept in the row of employees 201 equals the value of the same field in the row of departments 211. Constraint 309, finally, says that for all employees e1, all employees e2, and all departments, if the two employees are in the same department and employee e2 is the department manager, employee e1 's pay must be less than or equal to employee e2 's pay, or in other words, that no employee's pay may be higher than that of his manager.

Whenever an update operation is performed on relation 201, relation 211, or any on any view where the view is derived from relation 201 or 211 and the change involves a value of relation 201 or 211, a constraint check must be made to ensure that none of the constraints of FIG. 3 is violated. This constraint check is performed by update code 115. FIG. 4 shows how update code including the proper constraint check is generated. Constraints 301 and a transaction descriptor 403 which describes a class of transactions are input to update code generator 405, which generates update code 115 for the transaction described by descriptor 403. Update code 115 includes update constraints 407, which are checked whenever a transaction belonging to the class described by descriptor 403 is performed to make sure that the transaction is not violating one or more of the constraints 301.

FIG. 5 gives an example of how update code 115 is generated from constraints 301 and a transaction description 403. Transaction description 403 defines a class of hire transactions 501. In this transaction, done when a new employee is hired, a row 203 containing the new employee's name, salary, and department number is added to employees 201. The name of the new employee is represented in the transaction description by the parameter n, the salary by the parameter s, and the department number by the parameter dp. As indicated by the = symbol, the transaction description defines the hire transaction by equating it to a sequence of statements in a transaction language which describe the transaction.

In the example, only a single statement is required, namely an insert statement 503. Insert statement 503 specifies that a row 203 is to be inserted in employees 201, and that values are to be supplied for all three fields of the row. The order of the values corresponds to the order of the fields of the row. When update code generator 405 receives transaction description 403, it uses transaction description 403 and constraints 301 to generate update code 513. Update code 513 contains update constraint 515, which is an update constraint 407 which embodies the principles of the invention.

Inserting a row in employees 201 requires the use of the insert operation 509, one of the primitives 116 provided by data base management system 113. A property of insert operation 509 is that it aborts if the key of the row being added to the relation is the same as a key for a row which is already in the relation. Thus, if insert 503 is used to insert a row in relation 201 which has a value for n (representing the name) which is the same as a value for name field 205 in another record of employees 201, insert 509 will not alter employees 201.

In accordance with the principles of the invention, constraint 515 does not reflect the constraints which data base management system 113 requires for insert operation 509. Instead, update constraint 515 includes only those constraints which must be checked if insert operation 509 does not abort. Thus, none of the constraints in update constraint 515 deal with the relationship of the key 205 of the new row 203 to the keys 205 of the other rows 203 of employees 201. Update code generator 405 generates such a constraint 515 because insert statement 503 is one of a class of expressions in the language used for the transaction description whose evaluation requires that data base system 101 execute an operation which may abort. In FIG. 5, update constraint 515 is checked before the insert operation is performed; if the constraints of 515 are not satisfied, an error message 511 is returned by update code 115; if they are satisfied, insert operation 509 is executed, as indicated at the next to the last line. If insert operation 509 is successful, the new row is added to employees 201; if it aborts, portion 116 of data base management system 113 which handles primitive operations returns an error 119.

In more detail, update constraint 515 requires that the following be true before insert 509 is executed:

the salary specified for the new row is in the allowed range;

the department number specified in the new row is the same as that of one of the departments of relation 211; and

the salary specified for the new row is no more than that of the new employee's manager.

As may be seen by comparing the above constraints with the full set of constraints 301, update code generator 405 has transformed the full set of constraints into exactly those constraints which must be true before the specific transaction of adding a row to table 201 can be done.

4.3 Logical Description of the Technique

The following is a logical description of the technique of finding an update constraint in the presence of operations which may abort. The description begins with a definition of a simple transaction language which provides logical description of the transactions for which the weakest precondition is to be found. Of course, the technique is not limited to this language and may be applied with far more complex transaction languages.

4.3.1 Description of the Transaction Language

Simple updates μ are given by the grammar ##EQU1## and transactions B are defined to be sequences (blocks) of simple updates. ##EQU2## The abbreviations insert (r,a=t) and delete (r,b=t) will often be used. A simple update insert (r,a=t) is well-formed (with respect to the data base scheme) when a are the attributes of relation r and the terms of t contain no free record variables. If {κ1, . . . , κm } is the key for r, then delete (r,b1 =t1, . . . , bj =tj) is well-formed when {κ1, . . . , κm }.OR right.{b1, . . . , bj } and the terms of t contain no free record variables. A transaction B is well-formed if all of the simple updates comprising it are well-formed.

Transactions B(u), having all scheme variables among the variables u, will be referred to as transaction schemes. Given terms t=t1, . . . ,tm, an instance of such a scheme is written as B(t).

Informally, insert (r,a=t) inserts the record <a=t> into relation r and delete (r,b=t) deletes the the record τ in r with rb=t. The insert operation aborts if the insertion would result in duplicate keys, and the delete operation aborts if there is no record to be deleted. A sequence B aborts when any of its component updates aborts. The aborting semantics of the delete update is meant to model the semantics of the our current application. It is easy to modify our presentation to accommodate a delete that does not abort. In any case, we can formally define the semantics of transactions as a state-valued function B(u)ρs that takes a transaction scheme B(u), a state s, and a substitution p covering the variables u arguments.

Assume we have a transaction B, formulas α and β, and a substitution ρ that covers B, α and β. The formula β is a precondition of α w.r.t. B and ρ if for all database states s, s,ρβ, implies that Bρs,ρα--that is, if β holds before the execution of B, then α will hold afterwards. A formula β is called a weakest precondition of α w.r.t. B when β is a precondition of α w.r.t. B and for all database states s, if Bρs,ρα, then s,ρβ. In other words, if β is a weakest precondition of α w.r.t. B, then s,ρβ if and only if Bρs,ρα.

4.3.2 Definition of the Weakest Precondition

The invention defines a particular weakest precondition of a formula α which represents with respect to a transaction B, wpc(α, B). Ordinarily, we might expect that

wpc(α,μ; B)=wpc(wpc(α,B),μ)

reflecting a compositional semantics for sequencing. However, this definition is not consistent with our semantics since either μ or B may abort with the result that μ; B aborts with no change to the database. In the case of an aborting transaction the weakest precondition should simply be α, or some equivalent formula.

This can be expressed by defining a predicate (B) (pronounced B is nice) that is true if and only if B does not abort. We then define a compositional wpcn (α, B) for transactions B that do not abort. A weakest precondition can then be defined as

wpc(α,B)=((B)wpcn (α,B))((B)α),

which says that either B is nice and wpcn (α, B) holds or B will abort and leave the state unchanged.

The formulas (B) and wpcn (α,B) are defined as

1. (delete(r,b=t)=∃x Σr xb=t

2. (insert(r,κ1 =t1, . . . ,κm =tm, . . . ai =ti))=∀xΣr xκ1 ≠t1 . . . xκm ≠tm (where {κ1, . . . ,κm } is the key of r.)

3. (μ; B)=(μ)wpcs (b),μ)

where wpcs (α,μ) is a weakest precondition for simple updates, defined by recursion on the structure of α as ##EQU3##

4.4 Example Industrial Application of the Technique

The technique described above has been used to maintain the data integrity of an embedded relational data base in the 5ESS switch manufactured by AT&T Corp. (5ESS is a trademark of AT&T Corp.). The data base contains information concerning telephone customers, switch configuration, and network topology. Data integrity is crucial for the proper functioning of the switch since all application programs--performing tasks such as call processing, switch maintenance, and billing--are driven from information in the database.

The 5ESS database is large and complex, containing nearly one thousand relations. The switch developers had been dealing with the integrity problem in two ways: by continually auditing the data base and by writing "safe" transactions. The auditing was done by a program which ran continually and reported errors as it found them. The code for the program was generated using a formal language for declaring static integrity constraints--a variant of the tuple relational calculus with range restrictions. These integrity constraints involved tens of thousands of integrity assertions, some very complex.

A safe transaction is one which will never result in a violation of the integrity constraints. The process of writing safe transactions for the 5ESS switch required enormous effort and was prone to error. It involved the tasks of (1) writing an initial transaction in C using a low-level interface to a database manager, (2) locating and understanding the relevant integrity constraints, (3) determining which checks had to be made to ensure data integrity, (4) encoding these checks in C and inserting them into the transaction. Over a half million lines of transaction code have been handwritten in this way.

The techniques described above have been used to partially automate the process of writing safe transactions. We eased the burden of step (1) by replacing C with a high-level transaction language, and we replaced steps (2-4) with a code generator which automatically generates code for a safe transaction from a transaction description in the transaction language.

Our transaction language is based on Qian's first-order transactions, see Xiaolei Qian, "An Axiom System for Database Transactions", Information Processing Letters, 1990, vol. 36, pages 183-189. The first-order transactions include the basic operations of insert, delete, and update together with control constructs for sequencing, conditionals, and a restricted form of iteration. The code generated from the transaction language and the constraints for the switch has embodied the invention described above, and has thus efficiently specified safe transactions. Use of the transaction language has guaranteed correctness, has reduced the amount of code which the switch developers must write, and has reduced the time required to define a new transaction.

Parts A, B, and C of the Appendix of the present application provide an example of the application of our techniques in the 5ESS switch. Part A of the Appendix is the constraints which apply when a row is inserted into particular one of the relations in the data base. Part B is the transaction description for the operation of inserting the row into the relation. Part C is the insertion code generated from the constraints of part A and the transaction description of part B. It should be pointed out here that the insertion code has the same general form as the code in the example of FIG. 5; the update constraints are enclosed between verify and end verify and the primitive insert operations follow the update constraints, for example insert tRloscdb in Rloscdb in the ninth line of Part C.

The predicates discussed until now have had the relatively simple semantics of typical data base system insert operations. There is, however, no requirement of simplicity. Part D of the Appendix is a complex example predicate. The predicate is expressed in the declarative transaction description language used in the examples for the 5ESS switch and involves checks for key conflicts in two different relations as well as for legal indices in the second relation. If the checks are passed, the predicate specifies that a row is to be located in a third relation. The predicate of Appendix D can reflect the semantics of a system-level operation, as is the case with the predicates for the insert and delete operations in the data base system in the 5ESS switch, or it can be semantics which are expressly defined for a statement in the transaction description language. An example of a statement in the transaction description language for which semantics are expressly defined is the abort statement at the bottom of the first page of part B of the Appendix.

5 CONCLUSION

The foregoing Detailed Description has disclosed to those skilled in the art how update constraints may be employed to insure integrity in systems having operations which may abort and has shown how such update constraints may be employed in an industrial application. Use of the techniques is not limited to the data base systems disclosed herein. Rather, the techniques can be used in any application where a possible alteration is subject to constraints. For instance, the techniques might be used to verify that a modification to a computer program conforms to the requirements of the context in which it is made. It is furthermore apparent from the Detailed Description that the technique is applicable with many different ways of specifying constraints, transactions, and actual operations on the data base system.

All of the above being the case, the foregoing Detailed Description is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the law. ##SPC1##

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US4933848 *Jul 15, 1988Jun 12, 1990International Business Machines CorporationMethod for enforcing referential constraints in a database management system
US4947320 *Jul 15, 1988Aug 7, 1990International Business Machines CorporationMethod for referential constraint enforcement in a database management system
US5133068 *Aug 28, 1991Jul 21, 1992International Business Machines CorporationComplied objective referential constraints in a relational database having dual chain relationship descriptors linked in data record tables
US5267346 *Nov 14, 1991Nov 30, 1993Fujitsu LimitedCombination problem solving apparatus
US5355474 *Sep 27, 1991Oct 11, 1994Thuraisngham Bhavani MSystem for multilevel secure database management using a knowledge base with release-based and other security constraints for query, response and update modification
US5386557 *Nov 12, 1992Jan 31, 1995International Business Machines CorporationEnforcement of referential constraints in a database system
US5408657 *May 17, 1994Apr 18, 1995Unisys CorporationMethod of imposing multi-object constraints on data files in a data processing system
EP0351209A2 *Jul 12, 1989Jan 17, 1990International Business Machines CorporationMethod of performing operations in a relational data base management system
EP0428264A2 *Oct 8, 1990May 22, 1991International Business Machines CorporationA method for generating an access plan in a database system
Non-Patent Citations
Reference
1Anonymous, "Distributed Optimistic Concurrency control Based on Time-Stamp History Certification", IBM Technical Bulletin, Apr. 1991, vol. 33, No. 11, pp. 178-181.
2 *Anonymous, Distributed Optimistic Concurrency control Based on Time Stamp History Certification , IBM Technical Bulletin, Apr. 1991, vol. 33, No. 11, pp. 178 181.
3M. Lawley, R. Topor, M. Wallace, "Using Weakest Preconditions to Simplify Integrity Constraint Checking", Proceedings of the 4th Australian Database Conference, Brisbane, Feb. 1-2, 1993, pp. 161-170.
4 *M. Lawley, R. Topor, M. Wallace, Using Weakest Preconditions to Simplify Integrity Constraint Checking , Proceedings of the 4th Australian Database Conference, Brisbane, Feb. 1 2, 1993, pp. 161 170.
5 *M. Samg Gamal Eldino et al; Local and Global Constraints in Database Entegration IEEE, 1989, pp. 604 611.
6M. Samg Gamal-Eldino et al; "Local and Global Constraints in Database Entegration" IEEE, 1989, pp. 604-611.
7Marie Collins et al, "Security Constraint processing during the update operation in multijewel secure databare management system", IEEE, 1991, pp. 23-32.
8 *Marie Collins et al, Security Constraint processing during the update operation in multijewel secure databare management system , IEEE, 1991, pp. 23 32.
9P. W. P. J. Grefen, "Combining Theory and Practice in Integrity Control: A Declarative Approach to the Specification of a Transaction Modification Subsystem", Proceedings of the 19th VLDB Conference, Dublin, Ireland, 1993, pp. 581-589.
10 *P. W. P. J. Grefen, Combining Theory and Practice in Integrity Control: A Declarative Approach to the Specification of a Transaction Modification Subsystem , Proceedings of the 19th VLDB Conference, Dublin, Ireland, 1993, pp. 581 589.
11 *P. W. P. J. Grefen, Integrity Control in Relational Database Systems An Overview , Nov. 1991, Chapters 1, 2 and 3.
12P. W. P. J. Grefen, Integrity Control in Relational Database Systems-An Overview-, Nov. 1991, Chapters 1, 2 and 3.
13Sang Ho Lee et al, "Enforcement of integrity comstraints against transactions with transaction Axions," IEEE, 1992, pp. 162-167.
14 *Sang Ho Lee et al, Enforcement of integrity comstraints against transactions with transaction Axions, IEEE, 1992, pp. 162 167.
15Stephen Clarke et al, "Software fault trees and weakest preconditions: a comparasion and analysis", IEEE, 1993, pp. 225-236.
16 *Stephen Clarke et al, Software fault trees and weakest preconditions: a comparasion and analysis , IEEE, 1993, pp. 225 236.
17X. Qian, et al., "Integrity Constraint Reformulation for Efficient Validation", Proceedings of the 13th VLDB Conf., Brighton 1987, pp. 417-425.
18 *X. Qian, et al., Integrity Constraint Reformulation for Efficient Validation , Proceedings of the 13th VLDB Conf., Brighton 1987, pp. 417 425.
19Xiao Quan, "An Effective Method for Integrity Constraint Simplification" IEEE, 1988, pp. 338-345.
20 *Xiao Quan, An Effective Method for Integrity Constraint Simplification IEEE, 1988, pp. 338 345.
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US5752250 *Sep 12, 1995May 12, 1998Fujitsu LimitedInstance updating method and apparatus therefor
US5983231 *Feb 20, 1998Nov 9, 1999Fujitsu LimitedInstance updating method and apparatus therefor
US5987252 *Sep 19, 1997Nov 16, 1999Digital Equipment CorporationMethod and apparatus for statically analyzing a computer program for data dependencies
US6189010 *Jun 10, 1998Feb 13, 2001Platinum Technology, Inc.Method for repairing constraint violations in a database management system
US6304876Jun 5, 1998Oct 16, 2001Computer Associates Think, Inc.Method for enforcing integrity constraints in a database table using an index
US6408280Jul 22, 1999Jun 18, 2002Toshiba America Information Systems, Inc.Data driven constraints engine
US6427143 *Apr 10, 1998Jul 30, 2002Computer Associates Think, Inc.Method for loading rows into a database table while enforcing constraints
US6453314Jul 30, 1999Sep 17, 2002International Business Machines CorporationSystem and method for selective incremental deferred constraint processing after bulk loading data
US6820251 *Nov 6, 2000Nov 16, 2004Hewlett-Packard Development Company, L.P.System and method for a software recovery mechanism
US7069539 *Dec 10, 2001Jun 27, 2006Cisco Technology, IncApproach for implementing policy constraints in applications
US7299450 *Jun 17, 2003Nov 20, 2007Microsoft CorporationUndoing changes in a software configuration management system
US7315996Sep 22, 2005Jan 1, 2008International Business Machines CorporationMethod and system for performing heuristic constraint simplification
US7752006 *Jun 19, 2007Jul 6, 2010International Business Machines CorporationDevice, system and method of generating a hardware-verification test case
US7788616Nov 15, 2007Aug 31, 2010International Business Machines CorporationMethod and system for performing heuristic constraint simplification
US7793242Nov 15, 2007Sep 7, 2010International Business Machines CorporationMethod and system for performing heuristic constraint simplification
US8108367 *May 20, 2008Jan 31, 2012International Business Machines CorporationConstraints with hidden rows in a database
US8209301 *Oct 31, 2007Jun 26, 2012International Business Machines CorporationMethod and system for detection of integrity constraint violations
US20080059841 *Oct 31, 2007Mar 6, 2008International Business Machines CorporationMethod and system for detection of integrity constraint violations
Classifications
U.S. Classification717/106, 714/15, 717/122, 707/999.2, 714/47.1
International ClassificationG06F12/00, G06F9/46
Cooperative ClassificationG06F9/466
European ClassificationG06F9/46T
Legal Events
DateCodeEventDescription
Jun 17, 2008FPExpired due to failure to pay maintenance fee
Effective date: 20080430
Apr 30, 2008LAPSLapse for failure to pay maintenance fees
Nov 5, 2007REMIMaintenance fee reminder mailed
Sep 19, 2003FPAYFee payment
Year of fee payment: 8
Oct 4, 1999FPAYFee payment
Year of fee payment: 4
May 27, 1994ASAssignment
Owner name: AT&T CORP., NEW YORK
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:GRIFFIN, TIMOTHY G.;TRICKEY, HOWARD W.;REEL/FRAME:007028/0376;SIGNING DATES FROM 19940526 TO 19940527